International Journal of
Molecular Sciences
Article
Restricted Gene Flow for Gadus macrocephalus from
Yellow Sea Based on Microsatellite Markers:
Geographic Block of Tsushima Current
Na Song 1 , Ming Liu 1 , Takashi Yanagimoto 2 , Yasunori Sakurai 3 , Zhi-Qiang Han 4
and Tian-Xiang Gao 4, *
1
2
3
4
*
Fisheries College, Ocean University of China, Qingdao 266003, China; [email protected] (N.S.);
[email protected] (M.L.)
National Research Institute of Fisheries Science, Fisheries Research Agency, Yokohama 220-6115, Japan;
[email protected]
Graduate School of Fisheries Sciences, Hokkaido University, Hokkaido 041-8611, Japan;
[email protected]
Fishery College, Zhejiang Ocean University, Zhoushan 316022, China; [email protected]
Correspondence: [email protected]; Tel.: +86-580-255-0008; Fax: +86-580-208-9333
Academic Editor: Li Lin
Received: 8 January 2016; Accepted: 22 March 2016; Published: 29 March 2016
Abstract: The Pacific cod Gadus macrocephalus is a demersal, economically important fish in the family
Gadidae. Population genetic differentiation of Pacific cod was examined across its northwestern
Pacific range by screening variation of eight microsatellite loci in the present study. All four
populations exhibited high genetic diversity. Pairwise fixation index (Fst ) suggested a moderate to
high level of genetic differentiation among populations. Population of the Yellow Sea (YS) showed
higher genetic difference compared to the other three populations based on the results of pairwise
Fst , three-dimensional factorial correspondence analysis (3D-FCA) and STRUCTURE, which implied
restricted gene flow among them. Wilcoxon signed rank tests suggested no significant heterozygosity
excess and no recent genetic bottleneck events were detected. Microsatellite DNA is an effective
molecular marker for detecting the phylogeographic pattern of Pacific cod, and these Pacific cod
populations should be three management units.
Keywords: Pacific cod; microsatellite; genetic diversity; population genetic structure
1. Introduction
The Pacific cod Gadus macrocephalus is a demersal, economically important fish in the family
Gadidae. It is mainly distributed in the North Pacific, from the Yellow Sea of China through the Sea
of Japan, Okhotsk and Bering Seas to California in the eastern North Pacific [1,2]. As an important
cold-water species distributed in the Yellow Sea, the distribution of Pacific cod was closely related with
the Yellow Sea Cold Water Mass [3]. Low temperature provides a suitable environment for the survival
of Pacific cod and the Tsushima Current may delineate the distribution borders. The life history of
Pacific cod comprises demersal eggs, pelagic larvae and demersal adults [1,4,5]. The literature indicates
that this species could migrate for much longer distances than expected based on tag-recapture data,
and it can explain genetic homogeneity in Pacific cod over broad areas of the North Pacific [2,6]. Could
the Tsushima Current be a dispersal barrier of Pacific cod from the Yellow Sea? In the present study
we want to study how the Tsushima Current affects the dispersal of cold-water fish by checking the
genetic structure of Pacific cod.
Studies on population genetics have attracted much attention because such information is
important for understanding the pattern of biogeography and sustainable utilization of fishery
Int. J. Mol. Sci. 2016, 17, 467; doi:10.3390/ijms17040467
www.mdpi.com/journal/ijms
Int. J. Mol. Sci. 2016, 17, 467
2 of 11
resources [7]. If we ignore the population genetic structure, over-fishing may lead to the extinction
of the population with fewer individuals and reduce the genetic diversity of this species [8–10].
Until now, some genetic studies about the Pacific cod based on allozyme, mitochondrial DNA and
microsatellite DNA have been conducted; however, their results suggested different genetic diversity
and structures. Grant et al. [2] detected two major genetic groups (a western North Pacific Ocean
group and an eastern North Pacific group) by an allozyme marker. Moreover, small but significant
genetic differentiation was detected among northwestern Pacific populations based on mitochondrial
DNA [11]. However, Cunningham et al. [12] examined the genetic population structure of Pacific
cod by screening for variation at 11 microsatellite DNA loci, and higher genetic diversity across the
northeastern Pacific was found and genetic divergence highly correlated with geographic distance was
suggested. A strong genetic discontinuity between northwestern and northeastern Pacific populations
was detected by Canino et al. and the current distribution pattern represents groups previously isolated
during glaciations that are now in secondary contact [13].
These studies suggested different sensitivities of three molecular markers for detecting the genetic
structure and diversity of the Pacific cod. Compared with mitochondrial DNA and allozyme markers,
microsatellite DNA is a type of nuclear genetic marker that has been proven to be more sensitive
in detecting population genetic structure [14–16]. In the previous studies, only samples across the
northeastern Pacific range were collected [12], but no northwestern Pacific populations were employed.
In the present study, we collected fish samples from the Yellow Sea, the Sea of Japan, the Okhotsk Sea
and Eastern Hokkaido to conduct the genetic analysis and describe the genetic situation of Pacific cod
by comparing these results with the results of Cunningham et al. [12]. The results of the present study
will reveal the genetic structure and diversity of this economically important species and provide vital
information for sustainable exploitation and management of natural populations.
2. Results
Clear and unambiguous bands were identified by denaturing polyacrylamide gel electrophoresis.
All eight microsatellite markers were highly polymorphic and the number of alleles per locus varied
from eight to 24 (Table 1). The number of alleles (A), observed heterozygosity (Ho), expected
heterozygosity (He) and polymorphism information content (PIC) are shown in Table 1. The expected
heterozygosity (He) ranged from 0.496 (Gmo107, the Yellow Sea (YS)) to 0.957 (Gmo104, the Sea of
Japan (SJ)), and the observed heterozygosity ranged from 0.435 (Gmo102, SJ) to 1.000 (Gmo37, Eastern
Hokkaido (EH)). The PIC values ranged from 0.472–0.933, which suggested high genetic diversity of
this species (PIC > 0.5) [17]. Five deviations from Hardy-Weinberg equilibrium tests were detected after
Bonferroni correction: population EH at Gma107, population SJ at Gma102 and Gma104, population
the Okhotsk Sea (OS) at Gmo08 and population YS at Gma107, respectively. The null alleles for these
loci were also detected. Since the null alleles had a very low effect on the average genetic diversity of
our data, we retained all loci for further study. The test for linkage disequilibrium for populations and
loci showed a very low value of significant pairwise comparisons.
The values of Fst ranged from 0.0204–0.0618 and p-values were adjusted using the sequential
Bonferroni procedure (Table 2). Populations SJ and OS showed the lowest genetic divergence
(Fst = 0.0204; p < 0.05). Genetic differences between population YS and the other three populations were
larger and statistically significant after sequential Bonferroni correction. The Mantel test indicated no
significant relationship between pairwise estimates of Fst /(1 ´ Fst ) and geographic distance (r = 0.6366;
p = 0.132) (Figure 1). In order to further detect genetic structure, analysis of molecular variance
(AMOVA) was performed under three patterns of gene pools. Most of the variances were found within
populations for all three patterns. When two gene pools were tested, the values of FCT were largest but
not significant (FCT = 0.035, p > 0.05) (Table 3). The first and second principal components explained
73.26% of the overall variation in the three-dimensional factorial correspondence analysis (3D-FCA),
which separated all individuals into three groups (Figure 2).
Int. J. Mol. Sci. 2016, 17, 467
3 of 11
Table 1. Genetic diversity indices for Pacific cod.
Int. J. Mol. Sci. 2016, 17, 467
Location
Parameters Gma102
The
Yellow
Location
Sea (YS)
A
12
HE
0.892
Parameters
HO
A
12
Eastern
Hokkaido
(EH)
Eastern
Hokkaido
(EH)
0.861
Gmo19
Gmo37
12
12
0.91
0.496
Gma104
0.958
0.688
Locus
0.857
0.888
0.89
0.901
Gma107
0.583
Gma108
0.708
9
8
10
13
0.496
0.688
0.857
0.901
15
0.883
0.472
0.634
Gma109
0.762 Gmo08
0.826 Gmo19
0.8
0.817
0.871
12
0.853
0.583
0.958
0.583
0.708
0.762
0.826
HEPIC
0.869
0.861
0.922
0.883
0.728
0.472
0.594
0.634
0.878
0.817
0.893
0.871
0.852
0.853
HO A
0.667
12
0.917
18
0.565
10
0.625
11
180.917
14 0.875
120.875
PICHE
0.869
0.834
0.922
0.896
0.728
0.692
0.594
0.56
0.878
0.847
0.893
0.864
0.8520.82
HO
0.667
0.917
0.565
0.625
0.917
0.875
0.875
0.896
0.692
0.56
0.847
0.864
A
12
10
0.834
0.804
0.91
18
0.667
HE
0.804
PIC
0.758
HO
APIC
HE A
0.921
HO
14
10
0.857
0.921
0.714
0.862
0.89
0.812
0.659
10
0.621
0.682
11
0.758
24
0.89
13
0.812
14
0.621
0.5
0.836
18
15
0.915
0.888
14
12
0.7
0.883
0.7
0.88321
12
11
0.891
11
0.652
0.859
12
0.858
0.792
17
0.8580.91
17 1
0.910.881
1
12
0.881
0.901
12
0.783
0.891
0.827
Gmo37
0.792
0.89
0.82
12
0.915
0.8
0.859
15
0.682
0.659
0.714
0.836
PIC
0.862
10
0.857
0.871
0.5
PIC
10
11
0.667
0.871
11
HOHE
10
14
10
HO
The Sea of
Japan (SJ)
Gmo08
HO
A
A
The Sea of
Japan (SJ)
Gma109
0.892
HE
The
Okhotsk
Sea (OS)
Locus
Gma108
HE
PIC
The
Okhotsk
Sea (OS)
Gma107
8
13
Table15
1. Genetic 9diversity indices
for 10
Pacific cod.
Gma102
0.583
PIC
The Yellow
Sea (YS)
Gma104
0.859
0.95
0.901
0.867
0.652
0.783
0.95
0.827 14
0.859 13
0.86713
0.957
24
0.816
13
0.629
14
210.942
14 0.848
130.886
130.848
0.957
0.833
0.816
0.667
0.629
0.667
0.9420.87
0.848
0.833
0.886
0.833
0.848
0.875
0.833
0.667
0.667
0.87
0.833
0.833
0.875
0.933
0.78
0.602
0.917
0.814
0.854
0.817
0.933
0.78
0.602
0.917
0.814
0.854
0.817
3 of 11
Average
11.375
(2.264)
0.815
(0.148)
Average
0.752
(0.126)
11.375
0.781
(2.264)
(0.149)
0.815
(0.148) 14
0.752
(3.251)
(0.126)
0.831
0.781
(0.113)
(0.149)
0.805
14
(0.161)
(3.251)
0.8310.799
(0.113)
(0.115)
0.805
11.75
(0.161)
(1.909)
0.799
0.852
(0.115)
(0.087)
11.75
0.0751
(1.909)
0.852
(0.105)
(0.087)
0.815
0.0751
(0.089)
(0.105)
15.375
0.815
(4.565)
(0.089)
0.850
15.375
(0.101)
(4.565)
0.8500.760
(0.101)
(0.134)
0.760
0.819
(0.134)
(0.102)
0.819
(0.102)
A: allelic number; HO: observed heterozygosity; HE: expected heterozygosity; PIC: polymorphism
A: allelic number; HO : observed heterozygosity; HE : expected heterozygosity; PIC: polymorphism
information content.
information content.
2 distance (above diagonal) among populations of
Table
Table2.2.Pairwise
PairwiseFstFst(below
(belowdiagonal)
diagonal) and
and (δμ)
(δµ)2 distance
(above diagonal) among populations of
Pacific
cod.
Pacific
cod.
Population
Population
YS
YS
EH
EH
OS
OS
SJ
SJ
YS
–
–
0.0476 *
0.0476 *
0.0618
0.0618 **
0.0474 **
0.0474
YS
EH
EH
4.1303
4.1303
–
–
0.0328
0.0328 **
0.0249 **
0.0249
OS
OS
4.8946
4.8946
3.0133
3.0133
––
0.0204
0.0204 **
SJ
SJ
5.7433
5.7433
1.1149
1.1149
1.7213
1.7213
––
* Significantafter
after Bonferroni
Bonferroni correction.
* Significant
correction.
Figure
1. 1.Plot
of FFst/(1
/(1 −´FFst) )and
geographic distance. The reduced major axis
Figure
Plotofofpairwise
pairwise estimates
estimates of
st
st and geographic distance. The reduced major axis
(RMA)
regression
(RMA)
regressionline
lineoverlays
overlaysthe
the scatter
scatter plots.
plots.
Int. J. Mol. Sci. 2016, 17, 467
4 of 11
Int. J. Mol. Sci. 2016, 17, 467
Int. J. Mol. Sci. 2016, 17, 467
4 of 11
4 of 11
Table 3. The analysis of molecular variance of Pacific cod.
Table 3. The analysis of molecular variance of Pacific cod.
Table 3. The analysis of molecular variance of Pacific cod.
Gene Pools
Gene
Pools
Gene Pools
Structure Tested
Tested
Variance (%)
Statistics
Structure
Variance
(%) F Statistics
Structure Tested
Variance (%) F Statistics
(YS,
EH)
(YS,SJ,
SJ, OS,
––
––
(YS,
SJ, OS,
OS,EH)
EH)
–
–
One pool
Among populations
populations
3.86
FFST
One
pool
Among
3.86
ST = 0.039
0.039
One pool
Among
populations
3.86
FST =– 0.039
Within populations
populations
96.14
Within
96.14
–
Within populations
96.14
–
(YS)
vs.
(SJ,
OS,
EH)
–
–
(YS) vs. (SJ, OS, EH)
–
–
(YS)
vs.
(SJ,
OS,
EH)
–
Among groups
groups
3.45
FFCT
=–
0.035
Among
3.45
CT = 0.035
Two pools
Among
groups
3.45
0.035
Two pools
Within groups
2.07
FFSCCT==0.021
Two pools
Within
groups
2.07
FSC = 0.021
Within
populations
94.48
FFSTSC==0.055
Within
groups
2.07
0.021
Within populations
94.48
FST = 0.055
Within
populations
94.48
FST =– 0.055
(YS)
vs. (SJ,
EH) vs. (OS)
–
(YS) vs.
(SJ,
EH)
vs.
(OS)
–
Among
groups
1.09
FCT =––
0.011
(YS)
vs.
(SJ,
EH)
vs.
(OS)
–
Three pools
Among
groups
1.09
CT =
= 0.030
0.011
Within groups
2.95
FFSC
Among groups
1.09
FCT = 0.011
Three pools
Within
populations
95.97
FFST
Three pools
Within
groups
2.95
SC = 0.040
0.030
Within groups
2.95
FSC = 0.030
Within populations
95.97
FST = 0.040
Within populations
95.97
FST = 0.040
pp
p
––
–
0.00
0.00
0.00
––
–
––
–
0.23
0.23
0.23
0.00
0.00
0.00
0.00
0.00
0.00
–
–
0.49
–
0.49
0.74
0.49
0.00
0.74
0.74
0.00
0.00
Figure 2. Three-dimensional factorial correspondence analysis (3D-FCA) showing the relationships
Figure
Figure 2.
2. Three-dimensional
Three-dimensional factorial
factorial correspondence
correspondence analysis
analysis (3D-FCA)
(3D-FCA) showing
showing the
the relationships
relationships
among individuals of four populations based on eight microsatellite loci.
among
individuals
of
four
populations
based
on
eight
microsatellite
loci.
among individuals of four populations based on eight microsatellite loci.
The (δμ)222 genetic distance was calculated according to the allele frequency and the results
The (δµ)
(δμ) genetic
geneticdistance
distance
was
calculated
according
the allele
frequency
theshowed
results
The
was
calculated
according
to thetoallele
frequency
and theand
results
showed that population YS exhibited the larger genetic distance from the other three populations,
showed
that
population
YS
exhibited
the
larger
genetic
distance
from
the
other
three
populations,
that population YS exhibited the larger genetic distance from the other three populations, which
which was in accordance with the result of the pairwise Fst. However, 2the (δμ)22 genetic distance
which
was in accordance
theofresult
of the pairwise
Fst. However,
(δμ) distance
genetic between
distance
was
in accordance
with thewith
result
the pairwise
Fst . However,
the (δµ) the
genetic
between populations SJ and EH was the smallest, not
between populations SJ and OS. The topology
between populations
SJ was
and the
EH smallest,
was the smallest,
not between
populations
SJ and
The topology
populations
SJ and EH
not between
populations
SJ and2 OS.
TheOS.
topology
of the
of the unweighted pair-group method analysis (UPGMA) tree based on (δμ)
genetic distance showed
2
2
of the unweighted
pair-group
method
analysis
(UPGMA)
tree based
on (δμ)
genetic
distance
showed
unweighted
pair-group
method
analysis
(UPGMA)
tree based
on (δµ)
genetic
distance
showed
the
the relationship of the four populations more intuitively (Figure 3).
the relationship
of four
the four
populations
intuitively
(Figure
relationship
of the
populations
moremore
intuitively
(Figure
3). 3).
75
75
75
75
YS
YS
EH
EH
SJ
SJ
OS
OS
0.5
0.5
Figure 3. The unweighted pair-group method analysis (UPGMA) tree based on the (δμ)22 genetic
Figure 3.
3. The
unweighted pair-group
method analysis
analysis (UPGMA)
(UPGMA) tree
tree based
based on
on the
the (δµ)
(δμ)2 genetic
genetic
Figure
pair-group Results
method
distance
of The
four unweighted
Pacific cod populations.
of bootstrapping
(1000 replicates)
are shown
at nodes.
distance of
of four
four Pacific
Pacific cod
cod populations.
populations. Results
Results of
of bootstrapping
(1000 replicates)
replicates) are
are shown
shown at
at nodes.
nodes.
distance
bootstrapping (1000
Wilcoxon signed rank tests under the three mutational models (the infinite allele model (IAM);
Wilcoxon signed rank tests under the three mutational models (the infinite allele model (IAM);
Wilcoxon
signed
rank(SMM);
tests under
the three
mutational
models
(thewere
infinite
model which
(IAM);
stepwise
mutation
model
two-phase
mutation
model
(TPM))
notallele
significant,
stepwise mutation model (SMM); two-phase mutation model (TPM)) were not significant, which
stepwise mutation
model (SMM);
two-phaseexcess
mutation
(TPM))
significant,
suggested
that no significant
heterozygosity
was model
detected
(Tablewere
4). Innot
addition,
therewhich
is no
suggested that no significant heterozygosity excess was detected (Table 4). In addition, there is no
evidence for a significant deviation from the normal L-shaped distribution of allele frequencies as
evidence for a significant deviation from the normal L-shaped distribution of allele frequencies as
expected for a stable population under mutation-drift equilibrium.
expected for a stable population under mutation-drift equilibrium.
Int. J. Mol. Sci. 2016, 17, 467
5 of 11
suggested that no significant heterozygosity excess was detected (Table 4). In addition, there is no
evidence for a significant deviation from the normal L-shaped distribution of allele frequencies as
expected for a stable population under mutation-drift equilibrium.
Table 4. Results of Wilcoxon’s heterozygosity excess test, the mode shift indicator for a genetic
Int. J. Mol. Sci. 2016, 17, 467
5 of 11
bottleneck in four Pacific cod populations.
Table 4. Results of Wilcoxon’s heterozygosity excess test, the mode shift indicator for a genetic
Wilcoxon Sign-Rank Test
bottleneck in four Pacific
cod populations.
Population
Mode Shift a
IAM
TPM
SMM
Wilcoxon Sign‐Rank Test
0.125
0.844
0.990
L
Mode Shift a
IAM 0.680
TPM
1.000
1.000 SMM L
0.973
0.973 0.990 L
YS
0.125 0.125
0.844
L
1.000
1.000 1.000 L
EH
0.680 0.578
1.000
L
Numbers in the
table represent p-value;
distribution; IAM: L
the infinite allele
OS
0.125 a normal L-shaped
0.973allele frequency
0.973
model; SMM: stepwise
mutation
model;
mutation model.
SJ
0.578 TPM: two-phase
1.000
1.000
L
Population
YS
EH
OS
SJ
Numbers in the table represent p-value;
a
normal L-shaped allele frequency distribution; IAM: the
The Bayesian
implemented
in the
program
STRUCTURE
that the model
infinite allelealgorithm
model; SMM:
stepwise mutation
model;
TPM: two-phase
mutation indicated
model.
with K = 2 explained the data in a satisfactory manner, as this model resulted in the highest ∆K value
implemented
the program
that YS
the formed
model the
(Figures 4The
andBayesian
5). Twoalgorithm
clusters were
detectedinfrom
the fourSTRUCTURE
populations.indicated
Population
with K = 2 explained the data in a satisfactory manner, as this model resulted in the highest ΔK value
first cluster with 94.9% of the sampled individuals, and 91.6%, 90.4%, 91.9% from the other respective
(Figures 4 and 5). Two clusters were detected from the four populations. Population YS formed the
populations contributed to the second cluster (Table 5).
first cluster with 94.9% of the sampled individuals, and 91.6%, 90.4%, 91.9% from the other respective
populations contributed to the second cluster (Table 5).
Table 5. Proportion of eight populations in each of the two inferred clusters.
Table 5. Proportion of eight populations in each of the two inferred clusters.
Population
Population
YS
YS
EH
EHOS
OSSJ
SJ
Inferred Cluster
Inferred Cluster
1
2
1
2
0.949
0.051
0.949
0.051
0.084
0.916
0.084
0.916
0.096
0.904
0.096
0.904
0.081
0.919
0.081
0.919
Number of Individuals
Number of Individuals
24
24
24
2424
2424
24
Figure 4. STRUCTURE bar plots from eight microsatellite loci for four populations of Pacific cod.
Figure 4. STRUCTURE bar plots from eight microsatellite loci for four populations of Pacific cod. Black
Black lines separate individuals from different geographic areas. (a) K = 2; (b) K = 3; (c) K = 4; 1: YS;
lines separate individuals from different geographic areas. (a) K = 2; (b) K = 3; (c) K = 4; 1: YS; 2: EH;
2: EH; 3: OS and 4: SJ.
3: OS and 4: SJ.
Int. J. Mol. Sci. 2016, 17, 467
Int. J. Mol. Sci. 2016, 17, 467
6 of 11
6 of 11
Figure
5. The
Figure 5.
The model
model choice
choice criterion
criterion lnP(D)
lnP(D) for
for each
each KK value
value and
and graphical
graphical results
results of
of the
the
STRUCTURE
analysis.
STRUCTURE analysis.
3. Discussion
3. Discussion
The results of the present study showed that there were higher genetic differences between
The results of the present study showed that there were higher genetic differences between
population YS and the other three populations. The (δμ)22 genetic distance between population YS
population YS and the other three populations. The (δµ) genetic distance between population YS
and the other populations was larger than other pairwise genetic distances, which was supported by
and the other populations was larger than other pairwise genetic distances, which was supported
the results of 3D-FCA and STRUCTURE. In the marine environment, currents can be critical for
by the results of 3D-FCA and STRUCTURE. In the marine environment, currents can be critical
dispersal of marine fish, and then gene flow among populations may be influenced by marine
for dispersal of marine fish, and then gene flow among populations may be influenced by marine
currents [18,19]. Though tag-recapture data implied potential dispersal of Pacific cod, the Kuroshio
currents [18,19]. Though tag-recapture data implied potential dispersal of Pacific cod, the Kuroshio
Current and its branch, theTsushima Current, may block the gene exchange of Pacific cod because it
Current and its branch, theTsushima Current, may block the gene exchange of Pacific cod because
is a cold-temperature fish species. The optimal spawning temperature for Pacific cod ranges from 0
it is a cold-temperature fish species. The optimal spawning temperature for Pacific cod ranges from
to 13 °C,
but the warm-water currents such as the Tsushima Warm Current may act as an exchange
0 to 13 ˝ C, but the warm-water currents such as the Tsushima Warm Current may act as an exchange
barrier for this species due to their higher seawater temperature [20]. Population YS is enclosed by
barrier for this species due to their higher seawater temperature [20]. Population YS is enclosed by
land to the north and east, and by warm subtropical waters to the west and south [2,21]. The restricted
land to the north and east, and by warm subtropical waters to the west and south [2,21]. The restricted
gene flow between the Yellow Sea and the Sea of Japan was also supported based on allozyme
gene flow between the Yellow Sea and the Sea of Japan was also supported based on allozyme markers
markers by Gong et al. [22].
by Gong et al. [22].
The strong genetic differentiation between populations from the Sea of Japan and the Okhotsk
The strong genetic differentiation between populations from the Sea of Japan and the Okhotsk Sea
Sea determined with mitochondrial DNA by Liu et al. [11] was in contrast with the results of the
determined with mitochondrial DNA by Liu et al. [11] was in contrast with the results of the present
present study. Shaw et al. [23] examined the population genetic structure of the Patagonian toothfish
study. Shaw et al. [23] examined the population genetic structure of the Patagonian toothfish using
using mtDNA and nuclear DNA, and a similar situation was detected. He thought population
mtDNA and nuclear DNA, and a similar situation was detected. He thought population patterns may
patterns may reflect either genome population size effects or (putative) male-biased dispersal. The
reflect either genome population size effects or (putative) male-biased dispersal. The life history of
life history of Pacific cod comprises demersal eggs, pelagic larvae and demersal adults. Combined
Pacific cod comprises demersal eggs, pelagic larvae and demersal adults. Combined with the results
with the results of the present study and the life-history characteristics of Pacific cod, a smaller
of the present study and the life-history characteristics of Pacific cod, a smaller effective population
effective population size of mtDNA may enlarge the difference between populations and strong
size of mtDNA may enlarge the difference between populations and strong genetic differentiation
genetic differentiation was detected. In brief, a moderate to high level of genetic differentiation was
was detected. In brief, a moderate to high level of genetic differentiation was detected for Pacific cod
detected for Pacific cod in the present study, which suggested the high sensitivity of the microsatellite
in the present study, which suggested the high sensitivity of the microsatellite marker. The Bayesian
marker. The Bayesian clustering analysis by STRUCTURE suggested the Okhotsk Sea populations
clustering analysis by STRUCTURE suggested the Okhotsk Sea populations were genetically distinct
were genetically distinct from the populations of Japanese coastal waters when K = 3 was performed,
from the populations of Japanese coastal waters when K = 3 was performed, which was supported
which was supported
by the results of (δμ)2 genetic distance and 3D-FCA. The topology of the
by the results of (δµ)2 genetic distance and 3D-FCA. The topology of the UPGMA tree based on (δµ)2
UPGMA tree based on (δμ)2 genetic distance showed that populations from the Japanese coast
genetic distance showed that populations from the Japanese coast clustered with each other at first,
clustered with each other at first, and then with population OS. The pairwise Fst was calculated based
and then with population OS. The pairwise Fst was calculated based on the haplotype frequency, and
on the haplotype frequency, and genetic distance was based on the difference of alleles. There are
genetic distance was based on the difference of alleles. There are alleles that may cause this difference
alleles that may cause this difference between the two methods.
between the two methods.
The conservation of genetic diversity is very important for the long-term interest of any
The conservation of genetic diversity is very important for the long-term interest of any
species [24]. Studies showed that population size must be kept at a certain level because loss of
species [24]. Studies showed that population size must be kept at a certain level because loss of
heterozygosity could have a deleterious effect on population fitness [25]. In the present study, the
expected heterozygosity (He) of Pacific cod was higher than the average values of marine fish [26],
and most of the values of polymorphism information content (PIC) were more than 0.8, which
Int. J. Mol. Sci. 2016, 17, 467
7 of 11
heterozygosity could have a deleterious effect on population fitness [25]. In the present study, the
expected heterozygosity (He) of Pacific cod was higher than the average values of marine fish [26], and
most of the values of polymorphism information content (PIC) were more than 0.8, which suggested
that high genetic diversity was detected based on the microsatellite marker. The level of genetic
Int. J. was
Mol. Sci.
17, 467
7 of Pacific
11
diversity
in2016,
accordance
with the genetic examination of this species across its northeastern
range by Cunningham et al. [12]. Moreover, four populations indicated a similar genetic diversity and
suggested that high genetic diversity was detected based on the microsatellite marker. The level
did not show any geographical trends. Plenty of microsatellite analysis showed that marine fish usually
of genetic diversity was in accordance with the genetic examination of this species across its
exhibit
high genetic
diversity,
which
may be et
attributed
to the high
rate
of microsatellites
northeastern
Pacific
range by
Cunningham
al. [12]. Moreover,
fourmutation
populations
indicated
a similar or
the huge
population
size
of
marine
fish
[14–16,27].
genetic diversity and did not show any geographical trends. Plenty of microsatellite analysis showed
However,
of genetic
this species
was which
detected
mitochondrial
that marinelow
fishgenetic
usuallydiversity
exhibit high
diversity,
maybybeallozyme
attributedand
to the
high mutationDNA,
whichrate
was
contrast with
ourhuge
results
and Cunningham
al. [14–16,27].
[12]. Especially very low nucleotide
of in
microsatellites
or the
population
size of marineetfish
However,
low based
geneticon
diversity
of this species
detected
by allozyme
and mitochondrial
diversity was
detected
mitochondrial
DNA was
by Liu
et al. [11].
The allozyme
marker may be
which
in contrast with
ourasresults
Cunningham
et al. and
[12].the
Especially
very lowDNA
easilyDNA,
affected
by was
environmental
factors
it is a and
protein
level marker,
mitochondrial
nucleotide
diversity
was
detected
based
on
mitochondrial
DNA
by
Liu
et
al.
[11].
The
allozyme
was maternally inherited. These two markers may depart from neutral because they are easily
affected
marker may be easily affected by environmental factors as it is a protein level marker, and the
by selection pressure [28,29]. Previous studies showed that the common mutation rate of microsatellite
mitochondrial DNA was maternally inherited. These two markers may depart from neutral because
loci was faster than that of the mtDNA control region [13,30,31]. The results of the present study
they are easily affected by selection pressure [28,29]. Previous studies showed that the common
proved
that therate
microsatellite
marker
more than
sensitive
forthe
detecting
population
genetic diversity
mutation
of microsatellite
loci was
was faster
that of
mtDNA the
control
region [13,30,31].
The
of Pacific
cod.
results of the present study proved that the microsatellite marker was more sensitive for detecting
In
microsatellite
DNA cod.
is an effective molecular marker for detecting the
the conclusion,
population genetic
diversity of Pacific
In conclusion,
DNAThe
is results
an effective
markersuggested
for detecting
phylogeographic
patternmicrosatellite
of Pacific cod.
of themolecular
present study
thesethe
Pacific
phylogeographic
pattern
of
Pacific
cod.
The
results
of
the
present
study
suggested
these
Pacific
cod
cod populations should be considered as three management units and population YS should be paid
should of
be its
considered
as three management units and population YS should be paid
more populations
attention because
uniqueness.
more attention because of its uniqueness.
4. Experimental Section
4. Experimental Section
4.1. Fish Samples
4.1. Fish Samples
Samples from four locations across the northwestern Pacific Ocean were used in the present study
Samples from four locations across the northwestern Pacific Ocean were used in the present
(Figure 6, Table 6). A total of 96 individuals were collected during 2004–2005 and all individuals were
study (Figure 6, Table 6). A total of 96 individuals were collected during 2004–2005 and all individuals
identified
on the basis
of basis
morphology,
and a and
piece
of muscle
waswas
taken
from
each
were identified
on the
of morphology,
a piece
of muscle
taken
from
eachindividual
individual and
preserved
in
95%
ethanol
or
frozen
for
DNA
extraction.
and preserved in 95% ethanol or frozen for DNA extraction.
Figure 6. Sampling sites of Pacific cod in the present study. Shaded sea areas are continental shelves
Figure
6. would
Sampling
of Pacific
induring
the present
study.
are continental
shelves
that
havesites
been exposed
to cod
the air
periods
of lowShaded
sea-level.sea
KC:areas
Kuroshio
Current; YSWC:
that would
have
been
exposed
to
the
air
during
periods
of
low
sea-level.
KC:
Kuroshio
Current;
Yellow Sea Warm Current; SBCC: Subei Coastal Current; KCC: Korean Coastal Current; TSWC:
YSWC:
Yellow Warm
Sea Warm
Current;
SBCC: Subei
Coastal Current; KCC: Korean Coastal Current; TSWC:
Tsushima
Current;
OC: Oyashio
current.
Tsushima Warm Current; OC: Oyashio current.
Int. J. Mol. Sci. 2016, 17, 467
8 of 11
Table 6. Sampling sites, date of collection and sample size in the present study.
ID
Sampling Sites
Date
Number
YS
SJ
EH
OS
The Yellow Sea
The Sea of Japan
Eastern Hokkaido
The Okhotsk Sea
Total
January 2005
September 2005
January 2004
April 2004
–
24
24
24
24
96
4.2. DNA Extraction and PCR Amplification
Genomic DNA was isolated from muscle tissue by proteinase K digestion followed by a standard
phenol-chloroform method. Eight polymorphic microsatellite markers were employed for this
study [32,33] (Table 7).
Table 7. Information for eight microsatellite loci and primers of Pacific cod in the present study.
Locus
Repeat Motif
Primer Sequence (51 –31 )
(F, Forward; R, Reverse)
Size Range (bp)
GenBank No.
Gma102
(CTGT)17
F: TGGTTTCATTCGGTTTGGAT
R: GGGCTCAGGTAAAGCCTCTT
221–275
DQ027808
Gma104
(GA)8 (CAGAGACA)
(GAGACA)4 (GACA)16
F: AAAGAGAGCCACAGCCAGAT
R: ATTCAACTGTTGGCCTCTGC
168–230
DQ027810
Gma107
(CTGT)12
F: GGGAGTGGAGTACAGGGTGA
R: CCATTGTTTAACATCTGGGACA
195–243
DQ066622
Gma108
(GACA)7
F: AAGTCCCAACACACCAAAGC
R: CTCCTCTCTCGCGCTCTTTA
210–280
DQ027813
Gma109
(GTCT)7 G(GTCT)18
F: CATTTTACCTTTTGCTGAGGTG
R: AAATTAAATTAGTTAGATGGAAAGA
257–373
DQ066623
Gmo08
(GACA)
F: GCAAAACGAGATGCACAGACACC
R: TGGGGGAGGCATCTGTCATTCA
110–205
AFI159238
Gmo19
(GACA)
F: CACAGTGAAGTGAACCCACTG
R: GTCTTGCCTGTAAGTCAGCTTG
120–220
AFI159232
Gmo37
(GACA)
F: GGCCAATGTTTCATAACTCT
R: CGTGGGATACATGGGTACT
220–290
AFI159237
Polymerase chain reaction (PCR) was carried out in 10 µL volumes containing 0.5 U Taq DNA
polymerase (Takara Co., Dalian, China), 100 ng template DNA, 0.15 µM of each forward and reverse
primers, 0.2 mM each deoxy-ribonucleoside triphosphate (dNTPs), 10 mM Tris (pH 8.3), 50 mM KCl,
3.0 mM MgCl2 . The PCR amplification was carried out in a thermal cycler (Biometra, Göttingen,
Germany) under the following conditions: 5 min initial denaturation at 92 ˝ C, and then followed by
25 cycles of 30 s at 92 ˝ C, 30 s at annealing temperature 62–57 ˝ C, 30 s at 72 ˝ C and final extending for
30 min at 72 ˝ C. We used 8% non-denaturing vertical polyacrylamide gel electrophoresis to separated
PCR products and visualized by silver staining [34]. A sizing standard (100–300 base pairs) was run in
the center and at both ends of each gel to calibrate allele size. Furthermore, a reference sample was run
on each gel to ensure consistency in genotype scoring across runs.
4.3. Data Analysis
Deviations from the Hardy-Weinberg equilibrium (HWE) and linkage disequilibrium of each locus
within each site were checked by GENEPOP 3.4 [35]. The number of alleles (A), observed heterozygosity
(Ho), expected heterozygosity (He) and Polymorphism Information Content (PIC) were obtained by
Microsoft Excel tools [17]. The null alleles in each population were checked by Micro-Checker 2.2.3 [36].
The values of pairwise Fst were calculated by FSTAT 2.9.3 [37]. Bottleneck 1.2.02 program [38] was
used to detect the evidence of recent bottleneck events under three different mutation models: the
infinite allele model (IAM), stepwise mutation model (SMM) and two-phase mutation model (TPM),
Int. J. Mol. Sci. 2016, 17, 467
9 of 11
where 95% single-step mutations and 5% multiple steps mutations with 1000 simulation iterations
were set as recommended [39]. We also proposed a graphical descriptor of the shape of the allele
frequency distribution (mode shift indicator) which could differentiate between bottlenecked and
stable populations [40]. ARLEQUIN 3.1 was employed to assess the population structure by the
analysis of molecular variance (AMOVA) [41], and AMOVA was carried out with different gene pools.
In order to detect the suitable groups for AMOVA, Fst measures were subjected to multidimensional
scaling (MDS) and plotted in two dimensions (applied using SPSS 11.0 (SPSS Inc., Chicago, IL, USA),
data not shown). The relationship between genetic distances and geographic distances was assessed
using Reduced Major Axis (RMA) regression and Mantel tests using IBDWS [42,43]. We calculated
the (δµ)2 genetic distance by POPULATION 1.2 [44] and constructed the UPGMA tree based on the
(δµ)2 genetic distance by MEGA 5 [45]. Three-dimensional factorial correspondence analysis (3D-FCA)
was performed in GENETIX 4.05 to explore population divisions and relationships of Pacific cod,
independent from prior knowledge of their relationships [46]. The possibility of cryptic population
structure of Pacific cod was detected by STRUCTURE 2.2 [47]. Markov chain Monte Carlo (MCMC)
consisted of 100,000 burn-in iterations followed by 1,000,000 iterations. The simulated K values ranged
from 1 to 10 (total sites). Ten independent runs were implemented for each specific K-value in order to
verify the consistency of the results. The ad hoc estimated likelihood of K (∆K) was used to determine
the most likely number of populations (K) based on the rate of change in the log probability of the data
(Ln P(D)) [48].
Acknowledgments: We are grateful to Lu Liu and Ning Li for helping data analysis. This work was supported by
the National Programme on Global Change and Air-Sea Interaction (GASI-02-PAC-Ydspr), the Public Science
and Technology Research Funds Projects of Ocean (201405010, 201305043) and Japan Society for the Promotion of
Science Grant in Aid for Scientific Research (P04475).
Author Contributions: Na Song and Tian-Xiang Gao conceived and designed the experiments; Ming Liu
and Na Song performed the experiments; Na Song, Ming Liu and Zhi-Qiang Han analyzed the data;
Takashi Yanagimoto and Yasunori Sakurai contributed materials tools; Na Song and Tian-Xiang Gao wrote
the paper.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
Hart, J.L. Pacific fishes of Canada. Bull. Fish. Res. Board Can. 1973, 47, 180–730.
Grant, W.S.; Chang, I.Z.; Kobayashi, T.; Stahl, G. Lack of genetic stock discretion in Pacific cod (Gadus
macrocephalus). Can. J. Fish. Aquat. Sci. 1987, 44, 490–498. [CrossRef]
Tang, Q.S.; Ye, M.Z. Exploitation and Protection of Offshore Fishery Resources for Shandong Province; Agricultural
Press of China: Beijing, China, 1990; pp. 43–211.
Phillips, A.C.; Mason, J.C. A towed, self-adjusting sled sampler for demersal fish eggs and larvae. Fish. Res.
1986, 4, 235–242. [CrossRef]
Sakurai, Y.; Hattori, T. Reproductive behavior of Pacific cod in captivity. Fish. Sci. 1996, 62, 222–228.
Shimada, A.M.; Kimura, D.K. Seasonal movements of Pacific cod, Gadus macrocephalus, in the eastern Bering
Sea and adjacent waters based on tag-recapture data. Fish. Bull. 1994, 92, 800–816.
Tudela, S.; Garca-Marn, J.L.; Pla, C. Genetic structure of the European anchovy, Engraulis encrasicolus, in the
north-west Mediterranean. J. Exp. Mar. Biol. Ecol. 1999, 234, 95–109. [CrossRef]
Utter, F.M. Biochemical genetics and fishery management: An historical perspective. J. Fish Biol. 1991, 39,
1–20. [CrossRef]
Hutchings, J.A. Collapse and recovery of marine fishes. Nature 2000, 406, 882–885. [CrossRef] [PubMed]
Cook, R.M.; Sinclair, A.; Stefansson, G. Potential collapse of North Sea cod stocks. Nature 1997, 385, 521–522.
[CrossRef]
Liu, M.; Lu, Z.C.; Gao, T.X.; Yanagimoto, T.; Sakurai, Y. Remarkably low mtDNA control-region diversity and
shallow population structure in Pacific cod Gadus macrocephalus. J. Fish Biol. 2010, 77, 1071–1082. [CrossRef]
[PubMed]
Int. J. Mol. Sci. 2016, 17, 467
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
10 of 11
Cunningham, K.M.; Canino, M.F.; Spies, I.B.; Hauser, L. Genetic isolation by distance and localized fjord
population structure in Pacific cod (Gadus macrocephalus): Limited effective dispersal in the northeastern
Pacific Ocean. Can. J. Fish. Aquat. Sci. 2009, 66, 153–166. [CrossRef]
Canino, M.F.; Spies, I.B.; Cunningham, K.M.; Hauser, L.; Grant, W.S. Multiple ice-age refugia in Pacific cod,
Gadus macrocephalus. Mol. Ecol. 2010, 19, 4339–4351. [CrossRef] [PubMed]
Carlsson, J.; McDowell, J.R.; Diaz-Jaimes, P.; Carlsson, J.E.; Bole, S.B.; Gold, J.R.; Graves, J. Microsatellite and
mitochondrial DNA analyses of Atlantic bluefin tuna (Thunnus thynnus thynnus) population structure in the
Mediterranean Sea. Mol. Ecol. 2004, 13, 3345–3356. [CrossRef] [PubMed]
Li, Y.; Han, Z.; Song, N.; Gao, T.X. New evidence to genetic analysis of small yellow croaker (Larimichthys
polyactis) with continuous distribution in China. Biochem. Syst. Ecol. 2013, 50, 331–338. [CrossRef]
Cheng, J.; Yanagimoto, T.; Song, N.; Gao, T.X. Population genetic structure of chub mackerel Scomber japonicas
in the Northwestern Pacific inferred from microsatellite analysis. Mol. Biol. Rep. 2015, 42, 373–382. [CrossRef]
[PubMed]
Botstein, D.; White, R.L.; Skolnick, M.; Davis, R.W. Construct ion of a genetic linkage map in man using
restriction fragment length polymorphisms. Am. J. Hum. Genet. 1980, 32, 314–331. [PubMed]
Grant, W.S.; Bowen, B.W. Shallow population histories in deep evolutionary lineages of marine fishes:
Insights from sardines and anchovies and lessons for conservation. J. Hered. 1998, 89, 415–426. [CrossRef]
Palumbi, S.R. Genetic divergence, reproductive isolation, and marine speciation. Annu. Rev. Ecol. Syst. 1994,
25, 547–572. [CrossRef]
Alderdice, D.F.; Forrester, C.R. Effects of salinity, temperature, and dissolved oxygen on the early
development of Pacific cod (Gadus macrocephalus). J. Fish. Res. Board. Can. 1971, 28, 883–902. [CrossRef]
Zhang, C.I. Pacific cod of south Korean waters. Bull. Int. North Pacific Fish. Comm. 1984, 42, 116–129.
Gong, Y.; Park, Y.C.; Kim, S.S. Study of the management unit of fisheries resources by genetic method 1.
Genetic similarity of Pacific cod in the North Pacific. Bull. Natl. Fish. Res. Dev. Agency 1991, 45, 47–61.
Shaw, P.W.; Arkhipkin, A.I.; Al-Khairulla, H. Genetic structuring of Patagonian toothfish populations in the
Southwest Atlantic Ocean: The effect of the Antarctic Polar Front and deep-water troughs as barriers to
genetic exchange. Mol. Ecol. 2004, 13, 3293–3303. [CrossRef] [PubMed]
Falk, D.A.; Holsinger, K.E. Genetics and Conservation of Rare Plants; Oxford University Press: New York, NY,
USA, 1991.
Reed, D.H.; Frankham, R. Correlation between fitness and genetic diversity. Conserv. Biol. 2003, 17, 230–237.
[CrossRef]
DeWoody, J.A.; Avise, J.C. Microsatellite variation in marine, freshwater and anadromous fishes compared
with other animals. J. Fish Biol. 2000, 56, 461–473. [CrossRef]
Gonzalez, E.; Zardoya, R. Relative role of life-history traits and historical factors in shaping genetic population
structure of sardines (Sardina pilchardus). BMC Evol. Biol. 2007, 7, 1–12. [CrossRef] [PubMed]
Lemaire, C.; Allegrucci, G.; Naciri, M.; Bahri-Sfar, L.; Kara, H.; Bonhomme, F. Do discrepancies between
microsatellite and allozyme variation reveal differential selection between sea and lagoon in the sea bass
(Dicentrarchus labrax)? Mol. Ecol. 2000, 9, 457–467. [CrossRef] [PubMed]
Bazin, E.; Glémin, S.; Galtier, N. Population size does not influence mitochondrial genetic diversity in
animals. Science 2006, 312, 570–572. [CrossRef] [PubMed]
Lai, Y.; Sun, F. The relationship between microsatellite slippage mutation rate and the number of repeat units.
Mol. Biol. Evol. 2003, 20, 2123–2131. [CrossRef] [PubMed]
Liu, M.; Lin, L.S.; Gao, T.X.; Yanagimoto, T.; Sakurai, Y.; Grant, W.S. What maintains the central north Pacific
genetic discontinuity in Pacific herring? PLoS ONE 2012, 7, e50340. [CrossRef] [PubMed]
Canino, M.F.; Spies, I.B.; Hauser, L. Development and characterization of novel di-and tetranucleotide
microsatellite markers in Pacific cod (Gadus macrocephalus). Mol. Ecol. Notes 2005, 5, 908–910. [CrossRef]
Miller, K.M.; Le, K.D.; Beacham, T.D. Development of tri- and tetranucleotide repeat microsatellite loci in
Atlantic cod (Gadus morhua). Mol. Ecol. 2000, 9, 238–239. [CrossRef] [PubMed]
Bassam, B.J.; Caetano-Anolles, G.; Gresshoff, P.M. Fast andsensitive silver staining of DNA in polyacrylamide
gels. Anal. Biochem. 1991, 19, 680–683.
Raymond, M.; Rosset, F. GENEPOP: Population genetics software for exact test and ecumenism. J. Hered.
1995, 86, 248–249.
Int. J. Mol. Sci. 2016, 17, 467
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
11 of 11
Oosterhout, V.C.; Hutchinson, W.F.; Wills, D.P.; Shipley, P. MICRO-CHECKER: Software for identifying and
correcting genotyping errors in microsatellite data. Mol. Ecol. Notes 2004, 4, 535–538. [CrossRef]
Goudet, J. FSTAT (version 1.2): A computer program to calculate F-statistics. J. Hered. 1995, 86, 485–486.
Piry, S.; Luikart, G.; Cornuet, J.M. BOTTLENECK: A computer program for detecting recent reductions in
the effective population size using allele frequency data. J. Hered. 1999, 90, 502–503. [CrossRef]
Zeng, L.Y.; Cheng, Q.Q.; Chen, X.Y. Microsatellite analysis reveals the population structure and migration
patterns of Scomber japonicus (Scombridae) with continuous distribution in the East and South China Seas.
Biochem. Syst. Ecol. 2012, 42, 83–93. [CrossRef]
Luikart, G.; Allendorf, F.W.; Sherwin, B.; Cornuet, J.M. Distortion of allele frequency distributions provides a
test for recent population bottlenecks. J. Hered. 1998, 12, 238–247. [CrossRef]
Excoffier, L.; Laval, G.; Schneider, S. Arlequin version 3.0: An integrated software package for population
genetics data analysis. Evol. Bioinform. 2005, 1, 47–50.
Bohonak, A.J. IBD (isolation by distance): A program for analyses of isolation by distance. J. Hered. 2002, 93,
153–154. [CrossRef] [PubMed]
Jensen, J.L.; Bohonak, A.J.; Kelley, S.T. Isolation by distance, web service. BMC Genet. 2005, 6, 13. [CrossRef]
[PubMed]
Olivier, L.; Wolf, Y.I.; Koonin, E.V.; Aravind, L. The role of lineage-specific gene family expansion in the
evolution of eukaryotes. Genome Res. 2002, 12, 1048–1059.
Tamura, K.; Peterson, D.; Peterson, N.; Stecher, G.; Nei, M.; Kumar, S. MEGA5: Molecular evolutionary
genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods.
Mol. Biol. Evol. 2011, 28, 2731–2739. [CrossRef] [PubMed]
Belkir, K.; Borsa, P.; Chikhi, L.; Raufaste, N.; Bonhomme, F. GENETIX 4.05, Logiciel Sous Windows TM Pour La
Génétique Des Populations. Laboratoire génome, Populations, Interactions (in French); Université de Montpellier II:
Montpellier, France, 2004.
Pritchard, J.K.; Stephens, M.; Rosenberg, N.A.; Donnelly, P. Association mapping in structured populations.
Am. J. Hum. Genet. 2000, 67, 170–181. [CrossRef] [PubMed]
Evanno, G.; Regnaut, S.; Goudet, J. Detecting the number of clusters of individuals using the software
STRUCTURE: A simulation study. Mol. Ecol. 2005, 14, 2611–2620. [CrossRef] [PubMed]
© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons by Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).